On Friday, New York Governor Kathy Hochul signed something called the Responsible AI Safety and Education (Raise) Act, meant to, on one hand, establish an AI safety regime, and on another, troll Silicon Valley Republicans like Marc Andreessen who have been trying to dictate tech policy during the second Trump Administration.
This comes just days after President Trump sent out an executive order that ostensibly blocks states from regulating AI.
According to the new state law, AI companies with more than $500 million in annual revenue must draft, publish, and follow formalized sets of safety procedures aimed at preventing “critical harm,” and will have to report safety issues within 72 hours or be hit with fines, which makes it stricter than California’s SB 53, which gives companies 15 days to report safety issues.
About a week ago on December 11, the Trump executive order called “Ensuring a National Policy Framework for Artificial Intelligence,” framed AI as a federal priority and outlined something called an “AI Litigation Task Force” at the Department of Justice. This task force will ostensibly have the job of challenging state AI laws determined to be in violation of the federal program on AI (basically doing nothing) according to the attorney general.
Even if the executive order turns out to lack a strong legal foundation, tying state laws up in legislation is still a dreary prospect, but New York State has rushed headlong into that eventuality with this law.
In an explainer for Axios published Friday, legal experts talking to Maria Curi and Ashley Gold averred that Trump’s executive order relies on a strange reading of parts of the Constitution, such as the Dormant Commerce Clause, which is usually interpreted as an attempt to prevent states from writing self-dealing laws that are unfair to other states—not laws that are simply meant to fill a legal vacuum left by the federal government
